The long-term goal of Project 1 (C. Rooney and L. Wang) is to improve the expansion and persistence of adoptively transferred tumor-specific T cells using two novel approaches. (1) Incorporation of a third-generation chimeric antigen receptor (CAR) for GD2 containing intracellular signaling domains for CD28 and OX40 is hypothesized to enhance and sustain T- cell responses intratumorally, while providing resistance to inhibitory ligands elaborated within the tumor microenvironment. (ii) An extratumoral proliferative boost should be attained by engrafting the GD2.CAR onto T cells specific for the varicella-zoster virus (VZV), for which there exists a potent live-attenuated booster vaccine that increases the in vivo proliferation of VZV-specific T cells via their TCRs. These broad concepts will be explored in a clinical trial of GD2.CAR-engrafted VZV-specific T cells for the treatment of patients with advanced GD2-positive sarcomas (Aims 1 and 2). Because this will be a first-in-man study, the transduced T cells will also incorporate an inducible safety switch based on dimerization of the caspase-9 molecule, which was validated in a recent clinical trial against graft-vs.-host disease. Unfortunately, the youngest pediatric sarcoma patients will be VZV-negative, making it difficult to generate VZV-specific autologous T cells and to justify vaccination of these seronegative patients with a live-attenuated virus. Hence, Aim 3 will evaluate in a preclinical model a cellular vaccine against GD2 that should provide extratumoral stimulation via the CAR. This effort will rely on the JF neuroblastoma (NB) cell line, which has been extensively tested as a cellular vaccine for NB. GD2 is strongly expressed by JF cells and can be presented to GD2,CAR-positive T cells both directly by the JFNB cells and indirectly by local antigen-presenting cells. Genetic modification of JFNB cells to express cytokines, such as ILI 5 and GM-CSF, that enhance T-cell proliferation and recruit and activate dendritic cells should promote extratumoral proliferation of the CAR-positive T cells in ananalogous way to stimulation of the native T cell receptor by VZV. In part therefore our project uses concepts developed in Project 3 and provides potential ways of increasing the safety and efficacy of projects 2 and 3 (the iC9 safety gene) and Project 4 (the DNR for TGFbeta).